SEO Audit 1.3.4

[qubn]

qubn updated SEO Audit with a new update entry:

SEO Audit v1.2.0 - Thread Auto-Fix, Email Digest & More

SEO Audit v1.2.0 Update​

v1.2.0 adds runtime auto-fix for thread pages, a periodic digest email, and graceful handling of restricted HTTP environments.

---

New Features​

• Thread meta description auto-fix - generates fallback descriptions for threads with empty or very short first posts. Cleans BBCode artifacts that slip through XF's regex stripping, and extends short descriptions (<120 chars) with forum and board context...

Read the rest of this update entry...

 

[briansol]

bug report / false-positive.

1)
not all brackets are bbcode.



user simply put an edit in brackets.







2)
Clicking how to fix does nothing.




3) feature request.

I need to be able to drill down into these numbers. Having a list of 27,000 things to look at, and no where to go look for them to correct is not helpful.

click the number, it will take you to a list of those threads/etc that show that error so that I can assign moderators to correct.

 

[qubn]

qubn updated SEO Audit with a new update entry:

SEO Audit v1.3.0 - Thread JSON-LD Enrichment, Drill-Down & Bug Fixes

SEO Audit v1.3.0 Update​

v1.3.0 enriches thread structured data by extending XF core's output instead of injecting duplicates, adds drill-down navigation for large reports, and fixes several issues reported by users.

---

New Features​

• Thread JSON-LD enrichment - extends XF core's native DiscussionForumPosting via class extension to add commentCount, isPartOf (DiscussionForum with forum name and URL), and author fallback for...

Read the rest of this update entry...

 

[qubn]

bug report / false-positive.

1)
not all brackets are bbcode.

View attachment 335302

user simply put an edit in brackets.

View attachment 335301





2)
Clicking how to fix does nothing.

View attachment 335303


3) feature request.

I need to be able to drill down into these numbers. Having a list of 27,000 things to look at, and no where to go look for them to correct is not helpful.

click the number, it will take you to a list of those threads/etc that show that error so that I can assign moderators to correct.

View attachment 335304

Thanks for the report, really appreciate the screenshots. All three are addressed in v1.3.0 which just went up.

1. BBCode false positives
   • You're right, the detection was way too aggressive. It was matching anything in brackets as BBCode when most of it is just users typing normally. Fixed it so it only flags the actual XenForo BBCode tags (b, url, quote, img, etc.). Things like [EDIT], [SOLVED], dates in brackets, all good now. Won't flag those anymore.
2. "How to fix" link
   • Found the bug, there was a JS/CSS mismatch causing it to toggle backwards. Clicking it was essentially hiding something that was already hidden. Works properly now, click it and you'll see the recommendation text expand.
3. Drill-down
   • Totally fair ask, especially at 27k issues. Every number in the report is now a clickable link, issue type counts, severity counts, category counts, per-forum breakdown. Click any of them and it filters right to that subset. Also added a per-page selector (50/100/200/500) and you can now sort the full list by impact (severity × page views) so the biggest bang-for-buck fixes float to the top.

Let me know how it works for you.

 

[briansol]

upgraded, rescanned.

Homepage h1 certainly exists.





How to fix works now.







All 6 are the same thread. Certianly, there has to be 1000s of these in old mod threads and old trash can forum threads that bots can't see and mods won't open.






I'm reporting a lot of bbcode in meta tags/etc.
is this an XF bug? i feel like it should be stripped natively. also has a line break in it...





Seeing a lot of false positives here as well.



looking at this thread, its a data table.
lots of line breaks...

no bbcode




Lots more to explore. Thanks for the tool, I've fixed a few things already because of it.

 

[frpsucre]

There's also an error when the title is too long or contains emojis; it causes an error when scanning a report. I also mean that if the title has emojis, it generates an error when running a report; it doesn't complete.

 

[briansol]

On the main admin page, new stats bar


links to a dead page, due to &ampersand:


url should be
/admin.php?seo-audit/report&report_id=5

 

[akok]

and when scanning I have a constant error (sitemap.xml contains invalid XML and could not be parsed.) although the file is readable and I don’t see any errors visually

 

[akok]

Let's continue
"robots.txt contains a broad "Disallow: /" rule that may block all crawling."
in case the robots block is set to block some bots. I'm not sure if this is a bug, but it's worth checking in the description.
When checking for broken internal links, the /&#039 appendix appeared, which is why many links are marked as broken, even though the link is working and there is no &#039 in the page code.

---

The same error applies to external links.

 

[qubn]

qubn updated SEO Audit with a new update entry:

SEO Audit v1.3.1 - Bug Fixes & False Positive Reduction

SEO Audit v1.3.1 - Bug Fixes & False Positive Reduction​

v1.3.1 addresses the issues reported by @briansol, @frpsucre, and @akok after v1.3.0. The main focus is eliminating false positive BBCode detections on forums with legacy content and fixing a crash when scanning threads with emoji or very long titles.

---

Bug Fixes​

• False positive BBCode detection - plain-text brackets...

Read the rest of this update entry...

 

[qubn]

SEO Audit v1.3.1 is now available. This update directly addresses the issues reported by @briansol, @frpsucre, and @akok.

Fixes:

• BBCode false positives - tightened the v1.3.0 whitelist further. Single-letter tags like [s], [b], [i] no longer match words like [s]ome or [b]efore. Also fixes false positives from truncated meta descriptions breaking BBCode tag pairs.
• Homepage H1 - the check now fetches the actual rendered homepage HTML instead of guessing from a style property. If your H1 is there, it'll find it.
• Emoji/long titles - threads with emoji characters or very long titles no longer crash scans. Titles are safely truncated respecting UTF-8 boundaries.
• Dashboard widget link - fixed the broken URL encoding on the "View report" link in the admin statistics bar.
• Excessive formatting - data table posts with lots of newlines but no BBCode no longer trigger false positives.
• Non-indexed forums - zero-views and thin-thread checks are now skipped for staff-only/non-indexed forums.
• robots.txt - bot-specific Disallow: / rules no longer trigger the "broad Disallow" warning. Only User-agent: * blocks are flagged.
• Internal links - HTML entities (', &) in BBCode URLs are now decoded before checking, fixing false broken-link reports.
• Sitemap parsing - handles gzip-compressed sitemaps, UTF-8 BOM, and HTML redirects. Now shows the actual parse error and response snippet when flagging invalid XML.

 

[frpsucre]

Thanks for the update

 

[akok]

I can say that nothing has changed with the links.
We have an error

Internal link returns HTTP 404: https://safezone.cc/threads/problem...zhenij-posle-primenenija-combofix.2204/&#039;

in the thread https://safezone.cc/threads/combofix-rukovodstvo-po-primeneniyu.2773/ look in the console

<a href="https://safezone.cc/threads/problemy-s-kodirovkoj-dos-prilozhenij-posle-primenenija-combofix.2204/" class="link link--internal">Если у Вас проблемы с с кодировкой DOS-приложений после применения ComboFix</a>

and if the post is being edited, the extra apostrophes are also not visible.

Thank you

 

[qubn]

qubn updated SEO Audit with a new update entry:

SEO Audit v1.3.2 - URL Extraction Fix

SEO Audit v1.3.2 - URL Extraction Fix​

Fixes a remaining issue reported by @akok where internal and external links were flagged as broken due to a trailing quote being appended during URL extraction.

---

Bug Fix​

• Link checker trailing quote - the plain URL regex was re-matching inside already-parsed [URL='...'] BBCode blocks, capturing the closing quote as part of the URL. This caused valid links...

Read the rest of this update entry...

 

[qubn]

I can say that nothing has changed with the links.
We have an error


in the thread https://safezone.cc/threads/combofix-rukovodstvo-po-primeneniyu.2773/ look in the console

<a href="https://safezone.cc/threads/problemy-s-kodirovkoj-dos-prilozhenij-posle-primenenija-combofix.2204/" class="link link--internal">Если у Вас проблемы с с кодировкой DOS-приложений после применения ComboFix</a>

View attachment 335380

and if the post is being edited, the extra apostrophes are also not visible.

Thank you

v1.3.2 is up - this should fix it.

The issue was in the URL extraction regex. When a link uses [URL='https://...'] with single quotes, the plain URL regex was re-matching inside that block and capturing the closing quote as part of the URL. That trailing quote caused the 404 and showed as ' in the report.

The fix strips all [URL] blocks before the plain URL regex runs, so it can't happen anymore.

Upgrade and re-scan. If you still see it, run this and share the output so I can see the exact BBCode:

SELECT message FROM xf_post WHERE thread_id = 2773 AND message LIKE '%2204%' LIMIT 1;

 

[akok]

Everything is great, I don't see any more errors in the links.

 

[akok]

Of all the points, I'm concerned about the error "sitemap.xml contains invalid XML and could not be parsed." External services and search engines (except Google) process the sitemap successfully. How can I double-check?

 

[kontrabass]

Hi! I'm loving this so far! We got an F lol. Question: We have like 5000 threads with no meta description (going back 25 years), and they are pure crap threads with no content in the first post. Is there a way to mass delete these?

 

[qubn]

qubn updated SEO Audit with a new update entry:

SEO Audit v1.3.3 - Sitemap Parsing Fix

SEO Audit v1.3.3 - Sitemap Parsing Fix​

Fixes an issue reported by @akok where large sitemaps were incorrectly flagged as "invalid XML."

---

Bug Fix​

• Sitemap XML false positive on large sitemaps - the scanner used streamed reads capped at 512KB, which truncated large sitemaps mid-tag and caused false "invalid XML" errors. Now reads the full response body (capped at 10MB for safety). Tested against a 2.1MB...

Read the rest of this update entry...

 

[qubn]

Of all the points, I'm concerned about the error "sitemap.xml contains invalid XML and could not be parsed." External services and search engines (except Google) process the sitemap successfully. How can I double-check?

v1.3.3 should fix this. Your sitemap is 2.1MB (14,743 URLs), and the scanner was using streamed reads capped at 512KB, which truncated the XML mid-tag and caused the parse error. It now reads the full body. Upgrade and re-scan, the error should be gone.

Hi! I'm loving this so far! We got an F lol. Question: We have like 5000 threads with no meta description (going back 25 years), and they are pure crap threads with no content in the first post. Is there a way to mass delete these?

Glad it's useful! The addon detects those threads but doesn't delete content, that's a moderation decision.

Two things that can help right now:

1. Enable the thread meta description auto-fix under SEO Audit > Auto-Fix. It generates fallback descriptions at render time for threads with empty or short first posts, no manual editing needed. That alone should clear a big chunk of those 5,000 issues from your score.
2. For threads that are truly worthless, you can use XenForo's inline moderation in the forum view to select and delete batches.

Bulk actions from the scan report (e.g., select all "no meta description" issues and act on them) is on the roadmap for a future version.